Paul Gemmill Email & Phone Number

Irving, Tx, Us

Irving, Tx, Us

Collaborated in the architecture and development of a commercial product to convert in house Oracle databases to the AWS cloud. Tools and methodologies used included Java. C. BASH, PL/pgSQL, AWS, Cassandra, PostgreSQL, Oracle, and LInux.Installed and maintained Java and PostgreSQL development and test environments.Successfully designed and developed, and implemented a Java based system using PostgreSQL PL/pgSQL trigger functionality to implement Oracle partition interval processing in an AWS Aurora PostgreSQL declarative partition environment.Researched the possibility of developing a Java based system using Oracle system table, log, and schema information to detect, analyze, and convert Oracle data and tables that could be more efficiently processed in a Cassandra AWS cloud environment.Used DevOps techniques to automate the development and testing of the migration project, where feasible, using Java, BASH, and database scripts.

Oversaw key database administration functions including automated scripting, configuration of datasets and features, design/data modeling/schema implementation of tables and schemas for Cassandra and MemSQL databases and clusters, and performance analysis and tuning in a Datastax Cassandra, MemSQL, and Oracle environment.Successfully automated a collection of monthly cstats data and developed an on demand reporting system to show disk utilization and statistical trends for all disk file systems for a large 120 node Cassandra cluster with a total of 452 TB of occupied disk data. Development was done using Groovy and BASH.Used BASH, Groovy, and CQL to develop monitoring and analysis code for Cassandra logs and system table data.Created Java REST and JMX systems to obtains Cassandra performance metrics and operational data from Datastax OpsCenter and Cassandra JMX interfaces.Assisted in the implemented and debugging of complex stored procedures for the MemSQL databases.Automated MemSql SSH and MemSQL-Ops access through development of BASH and HTML code.

Collaborated on the development of machine learning to study and predict blood sugar levels in diabetics.Designed and assisted in the implementation of a format project infrastructure using Linux, Cassandra, Eclipse, Java, Groovy, R, and Python.Installed and maintained an eight node Cassandra cluster as a project data store.Designed and developed the methodology to extract features from the raw medical study data.

Responsible for Datastax Cassandra/NoSQL database administration, DevOps development, and Java development.Successfully added 60 new Cassandra/Spark nodes in three data centers to an existing 60 node Cassandra only cluster. This provided efficient extraction of data for a machine learning platform for a critical banking system.Reintroduced the Cassandra repair feature to successfully stabilize a major Cassandra cluster consisting of 120 nodes on 66 servers. The cluster went from several node and disk failures per week to an average of six weeks without a single node failure. Successfully upgraded Cassandra clusters from Datastax DSE 4.8.5 to DSE 5.0.11 (Cassandra 2.1 to Cassandra 3.0).Successfully resolved a severe Cassandra tombstone issue (some large SSTables were 91% tombstones) using targeted compactions to improve performance and reduce SSTables size significantly. This resulted in recovered disk space and lowered space usage costs.Used various Cassandra compaction, split, and cleanup techniques to reduce the size of a major Cassandra cluster by 20 TB. This extended the life of the disk configuration resulting in cost savings.Developed DevOps scripts in Groovy/BASH to automate the collection and analysis of Cassandra disk usage, SSTable size, and Tombstone distribution data. This enabled better management of the Cassandra disks and identified potential problem resolution and cost saving opportunities.Developed DevOps scripts in Bash to automate the application of maintenance to Cassandra clusters. This resulted in significant time and costs over applying maintenance manually.Added new technical features to the Portal/Reagent application using Java 8, Eclipse, TestNG, and Mockito. Also added a security/logon feature.

Responsible for the automated conversion of an existing RPG/DB2 system to a Java/Oracle system for a large governmental agency.Analyzed the existing RPG code and developed Groovy programs to implement lexical finite state machines and recursive descent parsers to parse and convert the code with a target of 80% automated conversion. The programs converted 84% of the RPG code automatically with individual programs ranging from 55% to 96% automatic conversion.Designed and documented a Manual Conversion Process to complete the conversion from RPG to Java for that code which was not converted by the automated process.Mentored two Java developers on the RPG to Java Manual Conversion Process. This included teaching them a working knowledge of RPG.Working together the rest of the code was converted successfully using the Java Manual Conversion Process within the specified time frame. This resulted in a major government contract being renewed for CGI.

Alpharetta, Georgia, Us

Responsible for enhancements to a Groovy/GRAILS/MySQL power monitoring system. Used Apache POI and Groovy/GRAILS to extract the information from power monitoring spreadshhet reports emailed by technicians and load it into the MySQL database. This resulted in a significant time savings and a very significant reduction in errors over having the data entered manually.Administered the MySQL database including installation, data modeling, schema implementation, and performance tuning.Successfully implemented Apache ActiveMQ as a GRAILS message broker. This resulted in increased efficiency in routing data between various components of the power monitoring management software system.

Responsible for ETL projects to load client historical data in a Groovy, Bash, PostgreSQL, AWS, Linux, Spring Batch environment. Successfully reduced the amount of time required to load a days worth of client historical data from 22-24 hours to 2-4 hours. This was done by reducing the existing 102 step Spring Batch job used for the ETL to 80 steps and implementing step wise refinement of the data.Created automated procedures to implement regular vacuuming techniques to the PostgreSQL database used for the ETL projects. This resulting in a significant cost saving in the AWS cloud environment.Successfully loaded the client historical data for two clients.

Developed software to load materials data into a PostgreSQL database, analyze the data to detect auxetic materials, and store the results in a Hadoop cluster for a research project. Over 1,100 auxetic materials were found. Development done with Groovy, R, FORTRAN, and Python in a Linux environment.For Southwest Airlines developed an interface between their C++ crew scheduling system and Java Blaze Adviser (a rules engine) to implement new federal crew scheduling rules. The application was written in Java and C++ with the Java Native Interface (JNI) and Google Protocol Buffers. The goal of verifying one million crew scheduling requests in five minutes was successfully met.Groovy/Grails architect/developer for a network monitoring and analysis proof of concept using Groovy, Grails, Oracle, JSF, HighCharts, and JQuery for a Comptramatics client.Developer for a large financial client of Comptramatics. Development environment used Java, Eclipse, Oracle, Jboss, Spring MVC, JSP, and JQuery. Development strategy was full Agile including Story Points, Groomings, Sprints and Retrospectives. Eight sprint cycles successfully completed with zero defects.

Responsible for collaboration in the development on a large Java state governmental web information management system using Java, Oracle, iceFaces, agile, and Linux. Successfully completed eleven agile sprint tasks primarily in the user interface and business logic tiers.Successfully imported an exported copy of the large client Oracle test database for use in the local development environment.

Richardson, Texas, Us

Collaborated on the development of a Java/PHP/Oracle accounting system in a Software As a Service (SAAS) environment. Successfully implemented four major and several minor system enhancements with only a single uncaught defect. All design and development project projects completed on time.Appointed Problem Manager. A majority of all problems were resolved within the time spans required by service level agreements.A number of tool and script enhancements were cited for excellence during performance review.

Ba, Physics

Mba, Statistics And Operations Research

Ms, Physics

Education record